PHYSICAL REVIEW B 97, 085303 (2018)

Cr doping induced negative transverse magnetoresistance in Cd3 As2 thin films

Yanwen Liu,1,2 Rajarshi Tiwari,3 Awadhesh Narayan,3,4,5 Zhao Jin,6 Xiang Yuan,1,2 Cheng Zhang,1,2 Feng Chen,1,2 Liang Li,6
Zhengcai Xia,6 Stefano Sanvito,3,* Peng Zhou,8,† and Faxian Xiu1,2,7,‡
1
State Key Laboratory of Surface Physics and Department of Physics, Fudan University, Shanghai 200433, China
2
Institute for Nanoelectronic Devices and Quantum Computing, Fudan University, Shanghai 200433, China
3
School of Physics, AMBER and CRANN Institute, Trinity College, Dublin 2, Ireland
4
Department of Physics, University of Illinois at Urbana-Champaign, Champaign, Illinois USA, 61801
5
Materials Theory, ETH Zurich, Wolfgang-Pauli-Strasse 27, CH 8093, Zurich, Switzerland
6
Wuhan National High Magnetic Field Center, Huazhong University of Science and Technology, Wuhan 430074, China
7
Collaborative Innovation Center of Advanced Microstructures, Nanjing 210093, China
8
State Key Laboratory of ASIC and System, Department of Microelectronics, Fudan University, Shanghai 200433, China

(Received 17 August 2017; published 9 February 2018)

The magnetoresistance of a material conveys various dynamic information about charge and spin carriers,
inspiring both fundamental studies in physics and practical applications such as magnetic sensors, data storage,
and spintronic devices. Magnetic impurities play a crucial role in the magnetoresistance as they induce exotic
states of matter such as the quantum anomalous Hall effect in topological insulators and tunable ferromagnetic
phases in dilute magnetic semiconductors. However, magnetically doped topological Dirac semimetals are hitherto
lacking. Here, we report a systematic study of Cr-doped Cd3 As2 thin films grown by molecular-beam epitaxy. With
the Cr doping, Cd3 As2 thin films exhibit unexpected negative transverse magnetoresistance and strong quantum
oscillations, bearing a trivial Berry’s phase and an enhanced effective mass. More importantly, with ionic gating the
magnetoresistance of Cr-doped Cd3 As2 thin films can be drastically tuned from negative to positive, demonstrating
the strong correlation between electrons and the localized spins of the Cr impurities, which we interpret through
the formation of magnetic polarons. Such a negative magnetoresistance under perpendicular magnetic field and
its gate tunability have not been observed previously in the Dirac semimetal Cd3 As2 . The Cr-induced topological
phase transition and the formation of magnetic polarons in Cd3 As2 provide insights into the magnetic interaction
in Dirac semimetals as well as their potential applications in spintronics.

DOI: 10.1103/PhysRevB.97.085303

I. INTRODUCTION antilocalization (WAL) and weak localization (WL) [6,7].

More importantly, a quantum anomalous Hall effect has been
It is well known that magnetic doping can effectively
achieved by means of magnetic doping and proper Fermi-level
manipulate the magnetoresistance (MR) of a material and can
engineering [8–13].
possibly give rise to magnetism [1–4]. One notable example
At the same time, the great success of graphene and topo-
is that of dilute magnetic semiconductors, which combine
logical insulators [14–17] has generated a tremendous interest
in the same material the advantages of semiconductors and
in topologically protected semimetals. Hereinto, the existence
ferromagnets [1,3]. For instance, Mn substitution in GaAs
of Dirac semimetals, which host three-dimensional Dirac
provides a compound, (Ga,Mn)As, that exhibits negative MR
fermions in the bulk state, has been theoretically predicted
and carrier-mediated ferromagnetism, and several other mag-
and experimentally verified [18–22]. Among them, Cd3 As2
netically doped semiconductors and oxides have demonstrated
was identified to be an ideal Dirac semimetal, showing good
novel spintronic functionalities [3,4]. Another important ex-
chemical stability [21], extremely high mobility [22–24], and
ample is magnetically doped topological insulators (TIs). The
chiral anomaly induced negative MR [25–28]. The robust topo-
topologically nontrivial ferromagnetic phase and the magnetic
logical protection in Cd3 As2 has also been proved in different
quantum-phase transition to a paramagnetic trivial state have
geometries like bulk [22–24], mircobelts [25,29], nanowires
been, respectively, demonstrated in different TI systems when
[26–28], and thin films [30], making this material conveniently
increasing the magnetic doping [5]. The interplay between
exploited. When either the time-reversal symmetry or the
topological protection and the breaking of the time-reversal
inversion symmetry is broken, the degenerate Dirac nodes
symmetry by the magnetic impurity results in intricate MR
in a Dirac semimetal will separate into Weyl nodes, result-
behavior in TI thin films, such as the competition between weak
ing in a Weyl semimetal phase. While inversion-symmetry-
broken Weyl semimetals were realized in noncentrosymmetric
monosphides from the TaAs family [31–33], stable time-
*
Sanvitos@tcd.ie reversal symmetry-broken Weyl semimetals without coupling
† to orbital motion and spin through Landau quantization and
pengzhou@fudan.edu.cn
‡
Faxian@fudan.edu.cn Zeeman splitting are still being sought. An external magnetic

2469-9950/2018/97(8)/085303(8) 085303-1 ©2018 American Physical Society

YANWEN LIU et al.
field or magnetic doping can implement the breaking of
the time-reversal symmetry. A few theoretical studies have
predicted a Ruderman-Kittel-Kasuya-Yosida interaction and a
Kondo effect among the magnetic impurities in both Dirac and
Weyl semimetals [34–37], although the experimental evidence
remains lacking.
Here we report a systematic study of the magnetotransport
properties of Cd3 As2 thin films incorporating different con-
centrations of Cr doping. Our high-quality Cd3 As2 thin films
were grown by molecular-beam epitaxy (MBE), and a negative
transverse MR is observed in all magnetically doped samples,
suggesting that the doping induces spin-dependent scattering
from magnetic polarons. The evolution of the Shubnikov–de
Haas quantum oscillations indicates that the band structure and
the topological properties experience a topological quantum-
phase transition from a nontrivial to a trivial phase, a phase
transition that is triggered by the Cr doping. Remarkably, the
MR can be electrically manipulated via ionic gating and the re-
sulting switching between positive and negative MR serves as a
signature of magnetic polaron formation in Cr-doped Cd3 As2 .
These findings are in a good agreement with density-functional
theory and Monte Carlo calculations. Our results represent
a step toward the manipulation of the topological Dirac
semimetal phase via magnetic doping, and pave the way for po-
tential spintronic applications of topological Dirac semimetals.
II. EXPERIMENTAL AND COMPUTATIONAL DETAILS
Cd3 As2 thin films were grown in a CREATEC molecular
beam epitaxy (MBE) system with base pressure less than
2 × 10−10 mbar. High-purity Cd (99.9999%), As (99.9999%),
and Cr (99.9999%) were evaporated by dual-filament, valve-
cracker, and single-filament effusion cells, respectively.
Freshly cleaved mica substrates were annealed at 350 °C
for 30 min to remove the molecule absorption. During the
growth process, the substrate temperature was kept at 225
°C. The entire growth was in situ monitored by the reflection
high-energy electron diffraction system. The crystal structure
was determined by x-ray diffraction (Bruker D8 Discovery,
Bruker Inc., Billerica, MA, USA).
The thin films were patterned into standard Hall bar
geometry manually. The solid electrolyte was produced as
follows: LiClO4 (Sigma-Aldrich) and poly(ethylene oxide)
(Mw = 100 000, Sigma-Aldrich) powders were mixed with
anhydrous methanol (Alfa Aesar). The solution was stirred
overnight at 60 °C and served as the electrolyte. After the
application of solid electrolyte, the device was kept at 350 K for
30 min in vacuum to remove the moisture before the transport
measurements.
The magnetotransport measurements were performed in a
Physical Property Measurement System by Quantum Design
with a magnetic field up to 9 T. A home-made measure-
ment system including lock-in amplifiers (Stanford Research
830) and Agilent 2912 source meters was used to acquire
experimental data. Pulsed magnetic-field measurements up to
60 T were performed at Wuhan National High Magnetic Field
Center.
Density-functional-theory calculations were carried out by
employing the Perdew-Burke-Ernzerhof form of the exchange-
correlation functional [38] as implemented in the QUANTUM
085303-2
PHYSICAL REVIEW B 97, 085303 (2018)
ESPRESSO code [39]. The rotationally invariant form of the
Hubbard U correction was used [40]. A plane-wave cutoff of
545 eV was employed and the Brillouin zone was sampled
using an 8 × 8 × 4k-point grid for the self-consistent calcula-
tions. The I41/acd space group with an 80-atom unit cell for
Cd3 As2 was used [41].
III. RESULTS AND DISCUSSION
A. Growth and temperature-dependent resistance
Cd3 As2 thin films were grown by MBE with different Cr
doping concentrations. From our previous study we know that
undoped Cd3 As2 thin films thinner than 50 nm will experience
a band-gap opening and become trivial insulators [30,42].
Therefore, we primarily focus on Cd3 As2 thin films with a
thickness of about 100 nm, which is thick enough to hold
the Dirac feature. The Cr concentration can be fine-tuned by
adjusting the Cr evaporation temperature during the growth
(as discussed in the above experimental and computational
details). The composition of as-grown thin films is determined
by energy-dispersive x-ray spectroscopy (EDX). It is found that
Cr replaces Cd and yields the compound Crx Cd3−x As2 [43].
The single crystallinity with the growth face of (112) plane
of the as-grown thin films is examined by x-ray diffraction
(XRD) [Fig. 1(a)]. Standard Hall bar devices were built from
the thin films. The temperature-dependent residual resistance
R(T)/R(300 K) reveals the effect of the Cr doping. For undoped
thin films, the resistance increases as the temperature decreases
in the high-temperature regime and saturates at low tempera-
tures. As the Cr concentration increases, however, the samples
begin to exhibit a low-temperature metallic behavior, similar
to what was reported for half metals [44].
B. MR behavior of the Cr-doped Cd3 As2 thin films
To better understand the effect of the doping, we have per-
formed temperature-dependent MR measurements of different
Cr-doped Cd3 As2 thin films. The most striking feature arising
from the Cr substitution is the doping induced negative MR.
In order to reveal the intrinsic effect of the magnetic doping,
several representative samples are selected for the detailed
study (Fig. 2). The undoped sample exhibits several prominent
features of MR behavior [Fig. 2(a)], distinct from the extremely
large MR ratio observed in the bulk counterpart [22,24].
Around zero field, there is a small dip in the MR, ascribed
to WAL, which is also observed in Cd3 As2 nanostructures
[26–29]. Such MR behavior is reminiscent of the WAL in
TIs and topological crystalline insulators [45,46]. According
to the recent theoretical predictions [47], the WAL in Dirac or
Weyl semimetals originates from quantum interference, giving
a positive MR near zero field. Such a WAL phenomenon has
been observed in TaAs [32,48] and regarded as a transport
signature of Dirac and Weyl semimetals, apart from the chiral
anomaly [26,28,32,47–49]. With increasing the magnetic field,
the positive MR caused by the classical cyclotron motion
prevails over the quantum interference effect and the quantum
oscillations are clearly identified. This suggests a high mobility
exceeding 2500 cm2 V−1 s−1 , a value that is consistent with
the measured Hall mobility of 2300 cm2 V−1 s−1 (Supplemental
Material, Fig. 2 [50]).
Cr DOPING INDUCED NEGATIVE TRANSVERSE … PHYSICAL REVIEW B 97, 085303 (2018)

FIG. 1. The temperature-dependent resistance of different Cr-doped Cd3 As2 thin films. (a) XRD spectra of as-grown thin films at different
doping concentration. (b) The temperature-dependent residual resistance of different samples. The heavily doped samples show metallic behavior
at low temperatures.

Just a tiny amount of Cr dopants (x < 0.05) suppresses
dramatically the positive MR and provides a strong negative
contribution [Fig. 2(b)]. The fact that the WAL feature is still
preserved at zero field strongly suggests that there is only a
negligible change in the topology of Cd3 As2 at a light magnetic
doping. However, the negative MR accompanied by the strong
quantum oscillations persists up to 9 T, implying another
hidden mechanism rather than WL, which usually contributes
near zero field [47].
In order to verify the negative MR, a high magnetic pulsed
field up to 55 T was applied to different magnetically doped
samples [Fig. 2(c)]. The lightly doped sample (x  0.1) shows
a negative MR below 10 T and then a largely suppressed
positive MR under higher magnetic field. The heavily doped
case, however, boosts the effect of the magnetic doping further.
The heavily doped one (x  0.2) exhibits a negative MR up to
∼30 T with strong quantum oscillations. For magnetic fields
larger than 30 T, all the magnetically doped thin films display

FIG. 2. MR behavior of different Cr-doped Cd3 As2 thin films. (a), (b) The MR behavior at different temperatures for the undoped and
lightly doped samples with the same thickness of 100 nm. (c) The MR behavior of different doping samples under ultrahigh magnetic pulsed
field. (d) The thickness-dependent MR behavior for the heavily doped samples at 2 K.

085303-3
YANWEN LIU et al. PHYSICAL REVIEW B 97, 085303 (2018)

FIG. 3. Analysis of quantum oscillations. (a) The critical field and phase-coherence length of the lightly doped sample. (b) The critical
field and phase-coherence length of the different Cr doping samples at 2 K. (c) The Berry phase of the different doping samples, indicating the
quantum-phase transition from nontrivial to trivial state. The inset is doping-dependent oscillation frequencies. The error bars were generated
from the linear fitting process in the Landau fan diagrams. (d) The effective mass of different samples, suggesting that the Cr doping enhances
the effective mass. The quantum lifetime is extracted from the high-field quantum oscillations.

a rapidly growing positive MR, indicating that the system qi qi

approaches the quantum limit as all the carriers are condensed
into the lowest Landau level [51–53]. Such MR behavior in
the quantum limit is confirmed by our following analysis of
the quantum oscillations and is consistent with what was also
reported in Cd3 As2 nanowires with low carrier density [26,27].
The thickness-dependent MR of heavily doped samples
shows that the dimensionality serves as an additional factor
influencing the magnetotransport. The MR of the thin films
with a thickness of 50, 100, and 200 nm was measured, as
shown in Fig. 2(d). The negative MR ratio of the 50 nm
sample at 9 T is about 25%, while for the thicker samples
of 100 and 200 nm it is 10 and 2%, respectively. The thinner
sample gives a larger negative MR, showing that with reducing
the dimensionality the transport charges have much stronger
interaction with the local magnetic moment [4].
C. Localization analysis
In order to gain a quantitative understanding of the com-
petition between WAL and the negative MR induced by the
Cr doping, we have extracted the phase-coherence length of
different Cr-doped Cd3 As2 thin films (Supplemental Material,
Sec. 2 and Supplemental Material, Figs. 4 and 5), following
the formula for 3D Dirac or Weyl semimetals [49]:
√ positions. For Dirac fermions, the linear plot of 1/B vs N
qi B + B
2 2 2
qi B B intercepts the N axis at 0.5, which implies a nontrivial π
σxx = C1 2
qi
+ C2 c2 .
Bc + B 2 Bc + B 2
Here the parameters C1 and C2 are positive for WL
and negative for WAL, and the critical field Bc is related
to the phase-coherence length φ according to Bc ∼ h̄/e2φ .
The phase-coherence length φ decreases from ∼150 to
∼60 nm with increasing Cr doping, directly suggesting the Cr-
induced suppression of the WAL [Fig. 3(a)]. The temperature-
dependent φ of the lightly doped sample displays an abnormal
evolution [Fig. 3(b)]. The φ saturates at low temperature,
similar to the case of dilute fluorinated graphene, having
spin-flip states induced by the local magnetic moment [54].
D. Quantum oscillations analysis
Quantum oscillations provide a wealth of information about
the band structure and the topological properties of materials.
The oscillation frequency SF is related to the area of the
Fermi surface and the Berry’s phase B is related to the
topological property, which can be extracted from Landau
fan diagrams according to the Lifshitz-Onsager quantization
rule [55]: SBF = N + 21 − 2πB − δ = N + γ , where N is the
Landau level index, SF is obtained from the slope of the
Landau fan diagram, and γ is the intercept. Conductivity σxx
is converted from ρxx and ρxy using σxx = ρxx /(ρxx 2
+ ρxy
2
).
Here, the integer indices N denote the σxx minima positions in
1/B, while the half-integer indices represent the σxx maxima
Berry’s phase, whereas a value of around zero represents a

085303-4
Cr DOPING INDUCED NEGATIVE TRANSVERSE …
FIG. 4. Gate modulation of the magnetoresistance of the heavily doped sample. (a) The temperature-dependent resistance of different carrier
density. (b) The Fermi energy as a function of carrier density. (c) The MR behavior with different carrier density. (d) Density of states of the
undoped and Cr-doped samples (the Cr concentration is ∼4%).
trivial Berry’s phase [23,56,57]. The measured SF of different
Cr-doped Cd3 As2 thin films increases from ∼17 to ∼37 T,
suggesting the doping induced larger Fermi surface [Fig. 3(c)].
The intercepts provide evidence for a Berry phase transition
from nontrivial (the undoped sample) to trivial (the heavily
doped sample), unambiguously showing the topological phase
transition [Fig. 3(c)].
Such a quantum topological phase transition is also con-
firmed by the effective mass of charge carriers in the thin
films. Following the Lifshitz-Kosevich formula [55,58,59], the
oscillation component Rxx could be described by
 
SF
Rx x ∝ RT RD cos 2π +γ ,
B
where RT and RD are the reduction factors accounting for
the phase-smearing effects due to temperature and scattering,
respectively. The effective mass m∗ can be estimated from
the temperature-dependent oscillation Rxx according to the
2π 2 kB T m∗ /h̄eB
temperature-smearing factor RT ∝ sinh(2π 2 k T m∗ /h̄eB) , where
B
kB is the Boltzmann’s constant, h̄ is the reduced Planck’s con-
stant (Supplemental Material, Fig. 6). The resulting effective
mass of carriers in different Cr-doped Cd3 As2 thin films can be
extracted [Fig. 3(d)]. It is clear that the effective mass increases
from 0.03 me to 0.12 me with high Cr concentration. Such a
heavy effective mass in the heavily doped samples deviates
PHYSICAL REVIEW B 97, 085303 (2018)
from the usual small mass in Dirac systems, reconfirming the
topological phase transition.
The carrier lifetime τ could be
∗
obtained from the Dingle fac-
tor RD ∼ e−D , where D = πm eBτ
. Because of the complicated
MR background in small magnetic field, we have calculated
the quantum lifetime from the high-field data of Fig. 2(f) for
better accuracy (Supplemental Material, Fig. 7). Uncharacter-
istically, the quantum lifetime becomes longer as we increase
the Cr concentration. The quantum lifetime indicates the small-
angle scattering that dephases the momentum eigenstates of
the carriers and is sensitive to all processes related to the
Landau level broadening [22,23]. In fact, this could be directly
witnessed from the much sharper quantum oscillations of
(1) the heavily doped samples than that of the undoped and
lightly doped ones. Scattering from the Cr impurities in turn
decreases the transport lifetime, which could be proved by
the declining trend of the Hall mobility [Fig. 3(d)]. As Cr
replaces the Cd ions, the spin-orbit interaction between Cd
and As is significantly reduced, resulting in an increase of
spin lifetime. Meanwhile, the local magnetic moment induced
by the Cr impurities results in a series of well-resolved
spin-polarized subbands under magnetic field [4], instead of
the degenerated Landau levels as in the undoped sample.
As a result, the scattering is reduced by the alignment of
spins through the external magnetic field, giving rise to the
enhanced quantum oscillations with relatively long quantum
lifetime.
085303-5
YANWEN LIU et al.
E. Gate-tunable MR behavior
In order to gain further insight into the mechanism leading to
the negative MR by Cr doping, ionic gating has been applied to
the heavily doped thin films. Previously, gate-tunable quantum
oscillations and ambipolar effect have been demonstrated
in undoped Cd3 As2 thin films [42]. As the carrier density
decreases, the resistance of the thin film increases, which is
consistent with the electron-dominated transport [Fig. 4(a)].
The Fermi energy extracted from the quantum oscillations
decreases from ∼100 to ∼20 meV, indicative of a good
tunability of the heavily doped thin films. Moreover, as the
Fermi level moves toward the bottom of the conduction band,
the MR turns from entirely negative to positive [Fig. 4(c)],
showing a similar behavior as that of the undoped sample
[Fig. 2(a)]. Consequently, the effect of Cr doping is weakened
gradually when the carrier density decreases, or when the
Fermi level moves away from the conduction band. This
indicates that the Cr doping generates an effective impurity
band inside the conduction band. At low temperatures, the
Fermi level lies inside this impurity band and the negative
MR plays the dominant role. When the Fermi level is shifted
away from the conduction band by the ionic gating, it leaves
this impurity band, leading to the suppression of the negative
MR (Supplemental Material, Fig. 10). This model is also
supported by our density functional theory calculations as
shown in Fig. 4(d). For undoped Cd3 As2 , the projected density
of states (DOS) shows a band gap of about 0.1 eV, which is
bridged by the Dirac cone (not displayed here) as observed in
the band structure. In contrast, in the heavily doped samples
[see lower panel of Fig. 4(d)], a band emerges in the DOS
at the bottom of the conduction band, consistent with our
experimental observations. Such a band originates from the p-d
interaction between the Cr dopants and the valence band of the
native Cd3 As2 crystal, and there is a significant spin splitting of
the bands. Such a mechanism is similar to other conventional
diluted magnetic semiconductors, such as (Ga,Mn)As [4].
It is now fundamental to explore the mechanism behind
the negative MR in our magnetically doped samples. From
our experiments we find that the dopant concentration, the
film thickness, and the carrier density can all affect the
amplitude of the negative MR. In conventional ferromagnetic
conductor, it is common to observe such negative MR due to
the spin-dependent scattering [2]. However, in our samples,
the ferromagnetism is not observed either in the transport
or the magnetization measurements, and the same picture is
confirmed by our Monte Carlo calculations (Supplemental
Material, Figs. 4 and 9 and Supplemental Material, Sec. 3).
According to the EDX analysis, the Cr impurities substitute
for Cd in the Cd3 As2 . As a result, Cr impurities provide both
local magnetic moments and hole doping. For the magnetism,
the effective lattice is the Cd sublattice on which the localized
spins of Cr lie. However, intrinsic defects in the thin films
provide additional electron doping that compensates for the
holes originating from Cr. Long-range magnetism can then
be potentially activated by the p-d coupling, which is indeed
strong as confirmed by our electronic structure analysis for
Cr-doped crystals. However, the peculiar structure of the
Cd3 As2 lattice prevents the formation of a robust ferromagnetic
ground state. The Cd sublattice presents little coordination (see
detailed discussions in Supplemental Material, Sec. 4) so that
085303-6
PHYSICAL REVIEW B 97, 085303 (2018)
the percolation threshold needed for long-range magnetism
is reached only at a high Cr concentration [60]. We have
estimated that such percolation threshold is approximately at
10% (10% of the Cd sites are substituted with Cr) for a fourth-
nearest-neighbor Cr-Cr exchange interaction, a concentration
far beyond what is present in our samples. Furthermore, even
beyond the percolation threshold the critical temperature is
driven by the long-distance exchange coupling, which we
calculate to be small (see Supplemental Material, Table I). As
such, even in the case percolation is reached through a spinodal
arrangement of the Cr ions [61] the expected TC remains low.
In brief, our theoretical analysis returns Cr-doped Cd3 As2 as
a diluted paramagnet with a strong p-d coupling, namely it is
similar to Mn-doped II-VI diluted semiconductors [62].
Monte Carlo simulations were applied to obtain the fer-
romagnetic transition temperature, TC , over the entire range
of Cr doping, where we use exchange parameters obtained
from our first-principles calculations. For all the four cases
from the first to the fourth neighbor, the long-range order
is not generated until the Cr dopant reaches a percolation
threshold at some critical concentrations of xc . Here, xc can
reduce to 0.3 as one goes to a longer-range exchange including
the fourth-neighbor exchange. As previously discussed, the
introduction of Cr dopants suppresses the WAL behavior
and generates a mechanism giving rise to the negative MR.
The interplay between quantum localization and the mag-
netism is reminiscent of the magnetic polarons scenario in
diluted semimetals and oxides [4,63,64]. Magnetic polarons
are formed by the interaction between the carrier spins with the
localized spins of the magnetic impurities [4]. Our electronic
structure calculations showing a significant spin split of the
valence band support the hypothesis of the magnetic polarons
formation. Such polarons are likely to have a short polaron
radius so that the percolation is not possible and no long-range
ferromagnetic order is developed. Yet, the interaction between
the transport electrons and the magnetic impurities remains
strong and it can profoundly affect the transport properties.
With the magnetic field, the spins between adjacent polarons
are aligned well along the field direction. Also, the random spin
of intermediate regions between magnetic polarons may be
aligned with the increasing magnetic field. As a result, the spin-
dependent scattering by the magnetic polarons is suppressed,
leading to a negative contribution to the magnetoresistance.
Such a scenario can be also made plausible by several pieces
of evidence from our transport data. It has been demonstrated
that the magnetic polaron is favored by reducing the dimension-
ality of systems and by increasing the exchange constant and
carrier effective mass [4]. In our experiments with the thinner
samples the negative MR is more pronounced, whereas in the
heavily doped 400-nm-thick sample, which is the 3D case, the
negative MR is almost undetectable (Supplemental Material,
Fig. 8 and Supplemental Material, Sec. 1). In thin films less
than 100 nm thick, however, the negative MR dominates the
transport, suggesting the formation of magnetic polarons in
a quasi-2D case with the reducing thickness. All the heavily
doped thin films have a large effective mass exceeding 0.1 me .
This is a fingerprint of the formation of magnetic polarons. The
second evidence is the carrier-density-dependent MR behavior
in the heavily doped sample. By decreasing the carrier density
through ionic gating, we observe the decrease of negative MR
Cr DOPING INDUCED NEGATIVE TRANSVERSE …
and finally a revival of the positive MR with the carrier density
of ∼2 × 1012 cm−2 . With low carrier density, the strength of
interaction between the carriers and magnetic impurities is
greatly suppressed, suppressing the magnetic polarons, similar
to the gate-tunable ferromagnetism in the MnGe quantum dots
[65]. Consequently, quantum interference between electrons
resumes the dominant role and the WAL emerges again.
IV. CONCLUSION
In conclusion, we have systemically studied the transport
properties of Cr-doped Cd3 As2 thin films. Magnetotransport
measurements were performed for the films with different
doping level, thickness, and carrier density. The magnetic
doping induced metallic behavior and negative MR were
observed at low temperatures. From the analysis of quan-
tum oscillations, we have demonstrated the doping induced
topological quantum-phase transition from nontrivial Dirac
semimetal to trivial dilute magnetic semimetal as the Cr doping
increases. For the lightly doped sample, the robust Dirac
feature is preserved, demonstrating that Cd3 As2 is a good
platform to study the magnetic interaction of Dirac fermions
[1] T. Dietl and H. Ohno, Rev. Mod. Phys. 86, 187 (2014).
[2] T. Dietl, J. Phys. Soc. Jpn. 77, 031005 (2008).
[3] T. Dietl, Nat. Mater. 9, 965 (2010).
[4] Introduction to the Physics of Diluted Magnetic Semiconductors,
edited by J. A. Gaj and J. Kossut (Springer, Berlin, 2010).
[5] J. Zhang, C.-Z. Chang, P. Tang, Z. Zhang, X. Feng, K. Li, L.
Wang, X. Chen, C. Liu, W. Duan, K. He, Q.-K. Xue, X. Ma, and
Y. Wang, Science 339, 1582 (2013).
[6] M. Liu, J. Zhang, C.-Z. Chang, Z. Zhang, X. Feng, K. Li, K. He,
L.-l. Wang, X. Chen, X. Dai, Z. Fang, Q.-K. Xue, X. Ma, and Y.
Wang, Phys. Rev. Lett. 108, 036805 (2012).
[7] Z. Zhang, X. Feng, M. Guo, K. Li, J. Zhang, Y. Ou, Y. Feng,
L. Wang, X. Chen, K. He, X. Ma, Q. Xue, and Y. Wang, Nat.
Commun. 5, 4915 (2014).
[8] C.-Z. Chang, J. Zhang, X. Feng, J. Shen, Z. Zhang, M. Guo, K.
Li, Y. Ou, P. Wei, L.-L. Wang, Z.-Q. Ji, Y. Feng, S. Ji, X. Chen, J.
Jia, X. Dai, Z. Fang, S.-C. Zhang, K. He, Y. Wang, L. Lu, X.-C.
Ma, and Q.-K. Xue, Science 340, 167 (2013).
[9] K. He, Y. Wang, and Q.-K. Xue, Natl. Sci. Rev. 1, 38 (2014).
[10] X. Kou, L. Pan, J. Wang, Y. Fan, E. S. Choi, W.-L. Lee, T. Nie, K.
Murata, Q. Shao, S.-C. Zhang, and K. L. Wang, Nat. Commun.
6, 8474 (2015).
[11] Y. Fan, P. Upadhyaya, X. Kou, M. Lang, S. Takei, Z. Wang, J.
Tang, L. He, L.-T. Chang, M. Montazeri, G. Yu, W. Jiang, T. Nie,
R. N. Schwartz, Y. Tserkovnyak, and K. L. Wang, Nat. Mater.
13, 699 (2014).
[12] X. Kou, Y. Fan, M. Lang, P. Upadhyaya, and K. L. Wang, Solid
State Commun. 215–216, 34 (2015).
[13] Y. Fan, X. Kou, P. Upadhyaya, Q. Shao, L. Pan, M. Lang, X.
Che, J. Tang, M. Montazeri, K. Murata, L.-T. Chang, M. Akyol,
G. Yu, T. Nie, K. L. Wong, J. Liu, Y. Wang, Y. Tserkovnyak, and
K. L. Wang, Nat. Nanotechnol. 11, 352 (2016).
[14] A. H. Castro Neto, N. M. R. Peres, K. S. Novoselov, and A. K.
Geim, Rev. Mod. Phys. 81, 109 (2009).
085303-7
PHYSICAL REVIEW B 97, 085303 (2018)
in 3D. In contrast, for the heavily doped samples the formation
of magnetic polarons is ascribed as the mechanism for the
negative MR. Our results not only provide a direction on the
research of Dirac semimetals, but also bring insights to the
potential spintronic applications based on topological Dirac
semimetals.
ACKNOWLEDGMENTS
This work was supported by the National Key Re-
search and Development Program of China (Grant No.
2017YFA0303302) and National Natural Science Founda-
tion of China (Grants No. 11474058 and No. 61674040).
F.X. acknowledges the support from the open project of
Wuhan National High Magnetic Field Center (Grant No.
PHMFF2015003). A.N. acknowledges support from ETH
Zurich. S.S. and R.T. acknowledge support from the European
Research Council (QUEST project) and from Science Foun-
dation Ireland (Grant No. 14/IA/2624). Part of the sample fab-
rication was performed at Fudan Nano-fabrication Laboratory.
Computational resources were provided by the Trinity Centre
for High Performance Computing (TCHPC).
[15] T. O. Wehling, A. M. Black-Schaffer, and A. V. Balatsky, Adv.
Phys. 63, 1 (2014).
[16] M. Z. Hasan and C. L. Kane, Rev. Mod. Phys. 82, 3045 (2010).
[17] X.-L. Qi and S.-C. Zhang, Rev. Mod. Phys. 83, 1057 (2011).
[18] Z. K. Liu, B. Zhou, Y. Zhang, Z. J. Wang, H. M. Weng, D.
Prabhakaran, S. K. Mo, Z. X. Shen, Z. Fang, X. Dai, Z. Hussain,
and Y. L. Chen, Science 343, 864 (2014).
[19] Z. Wang, H. Weng, Q. Wu, X. Dai, and Z. Fang, Phys. Rev. B
88, 125427 (2013).
[20] Z. Wang, Y. Sun, X.-Q. Chen, C. Franchini, G. Xu, H. Weng, X.
Dai, and Z. Fang, Phys. Rev. B 85, 195320 (2012).
[21] Z. K. Liu, J. Jiang, B. Zhou, Z. J. Wang, Y. Zhang, H. M. Weng,
D. Prabhakaran, S. K. Mo, H. Peng, P. Dudin, T. Kim, M. Hoesch,
Z. Fang, X. Dai, Z. X. Shen, D. L. Feng, Z. Hussain, and Y. L.
Chen, Nat. Mater. 13, 677 (2014).
[22] T. Liang, Q. Gibson, M. N. Ali, M. Liu, R. J. Cava, and N. P.
Ong, Nat. Mater. 14, 280 (2015).
[23] A. Narayanan, M. D. Watson, S. F. Blake, N. Bruyant, L. Drigo,
Y. L. Chen, D. Prabhakaran, B. Yan, C. Felser, T. Kong, P. C.
Canfield, and A. I. Coldea, Phys. Rev. Lett. 114, 117201 (2015).
[24] J. Cao, S. Liang, C. Zhang, Y. Liu, J. Huang, Z. Jin, Z.-G. Chen,
Z. Wang, Q. Wang, J. Zhao, S. Li, X. Dai, J. Zou, Z. Xia, L. Li,
and F. Xiu, Nat. Commun. 6, 7779 (2015).
[25] C. Zhang, E. Zhang, W. Wang, Y. Liu, Z.-G. Chen, S. Lu, S.
Liang, J. Cao, X. Yuan, L. Tang, Q. Li, C. Zhou, T. Gu, Y. Wu,
J. Zou, and F. Xiu, Nat. Commun. 8, 13741 (2017).
[26] H. Li, H. He, H.-Z. Lu, H. Zhang, H. Liu, R. Ma, Z. Fan, S.-Q.
Shen, and J. Wang, Nat. Commun. 7, 10301 (2016).
[27] C.-Z. Li, L.-X. Wang, H. Liu, J. Wang, Z.-M. Liao, and D.-P.
Yu, Nat. Commun. 6, 10137 (2015).
[28] L.-X. Wang, C.-Z. Li, D.-P. Yu, and Z.-M. Liao, Nat. Commun.
7, 10769 (2016).
[29] Z.-G. Chen, C. Zhang, Y. Zou, E. Zhang, L. Yang, M. Hong, F.
Xiu, and J. Zou, Nano Lett. 15, 5830 (2015).
YANWEN LIU et al.
[30] P. Cheng, C. Zhang, Y. Liu, X. Yuan, F. Song, Q. Sun, P. Zhou,
D. W. Zhang, and F. Xiu, New J. Phys. 18, 083003 (2016).
[31] C.-L. Zhang, B. Tong, Z. Yuan, Z. Lin, J. Wang, J. Zhang, C.-Y.
Xi, Z. Wang, S. Jia, and C. Zhang, Phys. Rev. B 94, 205120
(2016).
[32] C.-L. Zhang, S.-Y. Xu, I. Belopolski, Z. Yuan, Z. Lin, B. Tong,
G. Bian, N. Alidoust, C.-C. Lee, S.-M. Huang, T.-R. Chang,
G. Chang, C.-H. Hsu, H.-T. Jeng, M. Neupane, D. S. Sanchez,
H. Zheng, J. Wang, H. Lin, C. Zhang, H.-Z. Lu, S.-Q. Shen, T.
Neupert, M. Zahid Hasan, and S. Jia, Nat. Commun. 7, 10735
(2016).
[33] C.-L. Zhang, S.-Y. Xu, C. M. Wang, Z. Lin, Z. Z. Du, C. Guo,
C.-C. Lee, H. Lu, Y. Feng, S.-M. Huang, G. Chang, C.-H. Hsu,
H. Liu, H. Lin, L. Li, C. Zhang, J. Zhang, X.-C. Xie, T. Neupert,
M. Z. Hasan, H.-Z. Lu, J. Wang, and S. Jia, Nat. Phys. 13, 979
(2017).
[34] H.-R. Chang, J. Zhou, S.-X. Wang, W.-Y. Shan, and D. Xiao,
Phys. Rev. B 92, 241103 (2015).
[35] J.-H. Sun, D.-H. Xu, F.-C. Zhang, and Y. Zhou, Phys. Rev. B 92,
195124 (2015).
[36] A. K. Mitchell and L. Fritz, Phys. Rev. B 92, 121109 (2015).
[37] M. V. Hosseini and M. Askari, Phys. Rev. B 92, 224435 (2015).
[38] J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77,
3865 (1996).
[39] P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C.
Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I.
Dabo, A. D. Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R.
Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri,
L. Martin-Samos, N. Marzari, F. Mauri, R. Mazzarello, S.
Paolini, A. Pasquarello, L. Paulatto, C. Sbraccia, S. Scandolo,
G. Sclauzero, A. P. Seitsonen, A. Smogunov, P. Umari, and R.
M. Wentzcovitch, J. Phys.: Condens. Matter 21, 395502 (2009).
[40] S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys,
and A. P. Sutton, Phys. Rev. B 57, 1505 (1998).
[41] M. N. Ali, Q. Gibson, S. Jeon, B. B. Zhou, A. Yazdani, and R. J.
Cava, Inorg. Chem. 53, 4062 (2014).
[42] Y. Liu, C. Zhang, X. Yuan, T. Lei, C. Wang, D. Di Sante, A.
Narayan, L. He, S. Picozzi, S. Sanvito, R. Che, and F. Xiu, NPG
Asia Mater. 7, e221 (2015).
[43] X. Yuan, P. Cheng, L. Zhang, C. Zhang, J. Wang, Y. Liu, Q. Sun,
P. Zhou, D. W. Zhang, Z. Hu, X. Wan, H. Yan, Z. Li, and F. Xiu,
Nano Lett. 17, 2211 (2017).
[44] T. Guan, C. Lin, C. Yang, Y. Shi, C. Ren, Y. Li, H. Weng, X.
Dai, Z. Fang, S. Yan, and P. Xiong, Phys. Rev. Lett. 115, 087002
(2015).
085303-8
PHYSICAL REVIEW B 97, 085303 (2018)
[45] M. Lang, L. He, X. Kou, P. Upadhyaya, Y. Fan, H. Chu, Y. Jiang,
J. H. Bardarson, W. Jiang, E. S. Choi, Y. Wang, N.-C. Yeh, J.
Moore, and K. L. Wang, Nano Lett. 13, 48 (2013).
[46] C. Zhang, Y. Liu, X. Yuan, W. Wang, S. Liang, and F. Xiu, Nano
Lett. 15, 2161 (2015).
[47] H.-Z. Lu and S.-Q. Shen, Phys. Rev. B 92, 035203 (2015).
[48] X. Huang, L. Zhao, Y. Long, P. Wang, D. Chen, Z. Yang, H.
Liang, M. Xue, H. Weng, Z. Fang, X. Dai, and G. Chen, Phys.
Rev. X 5, 031023 (2015).
[49] H.-Z. Lu and S.-Q. Shen, Front. Phys. 12, 127201
(2017).
[50] See Supplemental Material at http://link.aps.org/supplemental/
10.1103/PhysRevB.97.085303 for detailed analysis of quantum
oscillations and theoretical calculations.
[51] A. A. Abrikosov, Phys. Rev. B 58, 2788 (1998).
[52] A. A. Abrikosov, Europhys. Lett. 49, 789 (2000).
[53] J. Hu and T. F. Rosenbaum, Nat. Mater. 7, 697 (2008).
[54] X. Hong, K. Zou, B. Wang, S.-H. Cheng, and J. Zhu, Phys. Rev.
Lett. 108, 226602 (2012).
[55] H. Murakawa, M. S. Bahramy, M. Tokunaga, Y. Kohama, C.
Bell, Y. Kaneko, N. Nagaosa, H. Y. Hwang, and Y. Tokura,
Science 342, 1490 (2013).
[56] J. Xiong, Y. Luo, Y. Khoo, S. Jia, R. J. Cava, and N. P. Ong,
Phys. Rev. B 86, 045314 (2012).
[57] Z. J. Xiang, D. Zhao, Z. Jin, C. Shang, L. K. Ma, G. J. Ye, B. Lei,
T. Wu, Z. C. Xia, and X. H. Chen, Phys. Rev. Lett. 115, 226401
(2015).
[58] D. Shoenberg, Magnetic Oscillations in Metals (Cambridge
University Press, Cambridge, England, 1984).
[59] L. Li, G. J. Ye, V. Tran, R. Fei, G. Chen, H. Wang, J. Wang,
K. Watanabe, T. Taniguchi, L. Yang, X. H. Chen, and Y. Zhang,
Nat. Nanotechnol. 10, 608 (2015).
[60] J. Osorio-Guillén, S. Lany, S. V. Barabash, and A. Zunger, Phys.
Rev. B 75, 184421 (2007).
[61] K. Sato, H. Katayama-Yoshida, and P. H. Dederichs, Jpn. J. Appl.
Phys. 44, L948 (2005).
[62] H. Saito, W. Zaets, S. Yamagata, Y. Suzuki, and K. Ando,
J. Appl. Phys. 91, 8085 (2002).
[63] J. M. D. Coey, M. Venkatesan, and C. B. Fitzgerald, Nat. Mater.
4, 173 (2005).
[64] A. H. MacDonald, P. Schiffer, and N. Samarth, Nat. Mater. 4,
195 (2005).
[65] F. Xiu, Y. Wang, J. Kim, P. Upadhyaya, Y. Zhou, X. Kou, W.
Han, R. K. Kawakami, J. Zou, and K. L. Wang, ACS Nano 4,
4948 (2010).